@connect-xyz/withdraw-js@0.52.0
A JavaScript SDK that enables frontend applications to seamlessly integrate with the Connect Withdraw product.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package is a browser widget loader with no install-time execution or local-system capabilities. Its fixed, user-triggered vendor module loading and failure telemetry are not concrete malicious behavior.
Decision evidence
public snapshot- The manifest has no lifecycle scripts and exports only the compiled SDK entrypoint.
- The SDK dynamically inserts a fixed Connect-hosted web-component module only when the consumer renders the widget.
- On script-load failure it posts diagnostic telemetry, including decoded JWT claim fields, to fixed Faro collector URLs.
- The declaration file contains a Vite development URL override, but the published JavaScript entrypoint uses fixed URL mappings instead.
- No install hook, shell execution, filesystem access, environment harvesting, or self-dependency is present.
- No credential value is sent by the telemetry code; it extracts limited JWT claim metadata for failure reporting.
- Remote loading and telemetry are user-invoked widget behavior with package-aligned endpoints.
Source & flagged code
5 flaggedA bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.
dist/index.d.tsView on unpkgThe declaration file contains a Vite development URL override, but the published JavaScript entrypoint uses fixed URL mappings instead.
dist/index.d.tsView on unpkg · L202Findings
Affected versions and remediation
This report applies to @connect-xyz/withdraw-js@0.52.0.
See all versions of @connect-xyz/withdraw-js for other recorded verdicts.
- Review the evidence and your use of @connect-xyz/withdraw-js@0.52.0 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.